WO2005030431A1

WO2005030431A1 - Method for machining a three-dimensional surface

Info

Publication number: WO2005030431A1
Application number: PCT/EP2004/010724
Authority: WO
Inventors: Raul Hess
Original assignee: Peguform Gmbh
Priority date: 2003-09-26
Filing date: 2004-09-24
Publication date: 2005-04-07
Also published as: EP1667812A1; DE10345081A1; US20070120842A1; ES2305827T3; ATE396007T1; EP1667812B1; PT1667812E; DE502004007246D1

Abstract

The invention relates to a method for removing material from a three-dimensional surface (1) of any shape in a multi-layered manner by means of a material removing agent (9), such as a laser, that acts in points on a surface. According to said method, a surface structure (2) is removed from the three-dimensional surface (1), the surface (1) is approximated by a polygon network (18), and raster images are associated with the polygons (19) of the polygon network (18). Said polygons are projected onto the machining region (10) of the material removing means. A series of superimposed polygon networks (18) is produced for the removal of the material, each of the polygon networks being associated with a layer in which material is removed, if the information stored on the projected raster image indicates the planned removal of material. The use of polygon networks (18) for removing material in a multi-layered manner ensures a very precise removal of material from a surface structure (2) of any shape, thus creating a highly precise three-dimensional surface structure (2).

Description

Process for processing a three-dimensional surface

The layer-by-layer removal of a material layer from a mold for the production of any three-dimensional surface structure has hitherto been carried out by means of etching processes or galvanic processes in which a positive mold with the desired surface design has been coated with a metal, which is then a negative mold for producing the desired molded part or the slide results. These process variants always require a large number of process steps in order to obtain a negative shape for only a single surface design. As a result, the same process steps are repeated each time the surface design is changed, which means additional costs and not inconsiderable expenditure of time in order to image surface structures with the precision such as that e.g. for the lifelike representation of a surface structure, such as a leather grain are required. So far, two methods have been widely used to economically scar tools, on the one hand, there is the etching scar, in which the surface of the workpiece is masked differently and then removed selectively by an etching liquid. This method can also be used in layers with restrictions and then creates a strongly graded transition between scar peaks and scar valleys. There are also difficulties with complicated geometries of the surface to be grained. Another method is the so-called galvano method. Here, a positive model, the so-called leather model, is covered with a film (or leather) which has the desired structure, for example leather grain. In a molding process, the scar is then transferred to a negative tool, which in turn is used to produce a (positive) bath model. A metal layer is then electroplated onto this in a bath. The electroplating tool obtained in this way then has to be reinforced, but can then only be used for certain methods for producing parts which do not stress its surface too much. The slush method and the spray skin method are particularly widespread. In addition, each of the latter methods is very time-consuming and costly.

Due to the great effort, which is associated with the processes known from the prior art and used on an industrial scale, there are approaches to the surface structure with a removal agent manufacture. A laser beam is a versatile removal agent. The technology of removing material by means of laser is known, for example, from DE3939866 A1 in the field of laser engraving. The removal of material by evaporation of a surface layer using a laser is known from DE4209933 C2. The laser beam is expanded and guided by rotatable deflecting mirrors over a reference line specified by a computer. The reference lines form a grid. The grid is scanned several times by the laser beam along angularly offset reference lines, material being removed by evaporation. By varying the direction of the laser tracks by rotating in the processing plane by a certain angle, systematic increases in the

Boundary layer avoided. This creates a network-like structure of the grid lines. This technology is only used on two-dimensional surfaces. A uniform removal of material in the grid is achieved by means of the technology disclosed in the patent specification. A line-by-line guidance of the laser in tracks (raster lines) or tracks in the respective processing field of the laser is disclosed in DE10032981 A1. The traces are applied in areas to a moving workpiece. In order to avoid that a sharp dividing line is formed in the overlap area of the tracks at the area boundaries, which is caused by excessive material removal in the overlap area, the area limits are offset with each removal. In other words, if an area is removed in a line, the laser does not stop at the edge along a line, but moves in the vicinity of this line. The end point of the removal is then in a distance range of this line, but this distance range differs from line to line. Since the end points are statistically distributed around the mean value of the line, no optical defect can be perceived. This method is suitable for removing grid fields that are on one level. As soon as the grid fields are inclined towards each other, a different amount of material is removed by the removal means when the removal means moves away from the grid field. Each individual end point would therefore have to be recorded, the material removal determined, and the material removal intended for the adjacent grid field corrected for the shortfall. For this reason, the method for three-dimensional surfaces can only be used with a high additional computing effort. [005] According to the teaching of DE10116672 A1, coarse and fine structures are processed differently, fine areas being processed by means of lasers and coarse areas using a lifting device. This technology is suitable in particular for processing metal surfaces which are arranged, for example, on printing cylinders. The rough machining is carried out by means of mechanical removal devices.

As is known, complicated structures can also be produced, for example, by material removal using a laser; this is used, for example, in the micromachining of materials. There are also processes for removing material over a large area with the laser.

It is an object of the invention, any surface structure, such as a leather grain, of arbitrarily designed three-dimensional

Remove surfaces. The object of the invention is therefore to develop a method which offers the possibility of providing tools and models of any shape with a three-dimensional surface structure which comes as close as possible to a natural or any other surface structure. Such a surface structure is, for example, the grain of the leather, which is characterized in that grain peaks have different heights and dimensions and the transition between grain peaks and grain valleys runs smoothly. It is a further object of the invention to avoid dividing lines or boundary lines when removing material. This object is achieved by the method according to claim 1, which can be used for different types of materials, can be carried out quickly in comparison to one of the methods known from the prior art, and no or little restrictions with regard to the geometries to be imaged of the surface structure. In addition, the method according to the invention is to be applied to any combination of materials. Single- or multi-layer material removal of an arbitrarily shaped three-dimensional surface is carried out by means of a removal agent acting point-wise on a surface, such as a laser, in which a surface structure is produced on the three-dimensional surface, the surface being approximated by at least one polygon mesh, each polygon of the Polygon mesh is assigned to the processing area of the laser. The surface approximated by the polygon mesh is scanned using a scanning device. The scanning device, for example a galvanoscanner, defines the processing area of the laser. The original three-dimensional computer model or master model of the workpiece is described by a sufficiently close-meshed polygon mesh, which in turn is derived from the CAD (spline) description of the workpiece. The three-dimensional corners of the polygons correspond to two-dimensional points in one or more original texture bitmaps, as a result of which the polygons are transferred into the two-dimensional space of the bitmap. The grayscale value of the bitmap corresponds to the required surface removal on the workpiece. Subsequently, processing areas for the individual layers are defined. The grayscale bitmaps for the polygons of the individual layers result from a parallel projection of the polygons and bitmaps of the original model onto the polygon of the processing area. The surface structure is thus described by at least one raster image, the processing area of the surface to be processed in each case coming completely into the focus area of the laser. The point position of the polygon corners in three-dimensional space corresponds to a two-dimensional coordinate position on the surface of the raster images. Material removal can take place in several layers, each layer being assigned its own polygon mesh. In an advantageous embodiment, the section of each layer to be processed does not have an edge section in common with one of the previously processed sections of another layer. The method for the layer-by-layer selective removal of shape on a workpiece aims to introduce a structure, for example in the form of a leather grain in the workpiece, which is characterized in that the transitions between scar peaks and scar valleys run as evenly as possible. Furthermore, as few restrictions as possible should be necessary with regard to the topology of the workpiece, ie there is no restriction here, for example, to cylinder surfaces as in the prior art, see, for example, DE 101 16672 A1. According to DE 4209933 C2, the structure is introduced, for example, by evaporating the material by means of a laser beam. This is guided computer-controlled along predetermined grid lines over the workpiece. In the case of large areas, processing generally takes place in sections (cf. also DE10032981 A1). Any surfaces and scars must be represented in such a way that they can be produced using a known method for removing material, in particular a laser method. A distinction must be made here between the description of the topology, ie the geometry of the workpiece and the scar, that is to say the desired fine structuring of the surface which is produced with the workpiece by a shaping process. In the automotive industry, NURBS (non-uniform rational B-splines) are generally used to describe the topology of so-called free-form surfaces. Since a single NURBS surface cannot describe a complex geometry satisfactorily, several so-called NURBS patches are put together. Often, these are also trimmed before assembly, for which purpose NURBS curves lying on the NURBS surfaces are used.

To be able to process this topology with the laser, it must be divided into processing areas. The size of the processing area is ideally selected so that it can be scanned only by influencing the galvanomirror when the scanner is in the appropriate position (approximately vertically on the processing area if possible). Furthermore, the change in distance between scanner and processing area should be kept small. When choosing the size of the machining area, the goal must always be that neither the angular position of the laser nor the change in the distance between the surface and the scanner results in an undesirable change in the thickness of the material removal or the material removed per unit of time. With each processing area, it should be noted that the laser as a whole focuses on it.

The possible processing area at a certain position of the scanner can be described by using the focus cuboid when using a plane field lens. The distance between the scanner and the center plane of the cuboid is given by the focal length of the laser optics. The height of the processing area, given the maximum error of the removed layer thickness, is given by the maximum depth of focus (= deviation from the focal length) and its side length by the corresponding maximum deflection of the galvanomirror in the scanner. Within the focus cuboid, the processing section can be approximated by a polygon, the corners of which all lie on a surface that ideally has the exact distance of the focal length from the laser optics and is perpendicular to the direction of the laser beam in the middle position of the deflection mirror. A surface section of the surface to be processed corresponds to this polygon, which is created by projecting the polygon onto the NURBS surface and must lie completely in the focus cuboid.

The entire topology of the surface to be processed is thus described by a grid of connected polygons of different sizes and shapes. The polygon edges are to be selected independently of the edges of the NURBS patches describing the surface to be processed, ie it can and it will happen that one or more points of the polygon lie on a patch and one or more points of the polygon on the adjacent NURBS patch. For the description of the fine structure of the surface, a raster image (bitmap) is assigned to each polygon for the purpose of better processability by the control program of the laser. The size of the pixel corresponds minimally to the size of the diameter of the laser light spot and the gray level (brightness) or the color level (intensity) of the pixel corresponds to the depth of the structure at this point. For example, a white dot would mean that no material would be removed at all, while a black dot would mean maximum material removal (or vice versa).

An even higher accuracy can be achieved by a description of the laser point by several pixels in the bitmap. The disadvantage is the enlargement of the bitmap and the correspondingly higher memory requirement and computing effort in the control electronics. The coding of the bitmap corresponds to the maximum number of layers, that is to say 256 layers of gray (= 8 bits) per pixel, a maximum of 256 layers can be displayed. Various computer formats with corresponding compression algorithms are known for storing this raster image, which result in a very large reduction in the memory requirement. In the general case, the polygon will rarely have a square shape. Therefore, the corner points of the polygon in three-dimensional space are assigned to a corresponding point on the bitmap in 2 D coordinates (texture coordinates). If the polygons are arranged accordingly, it is also possible to combine the texture coordinates of several polygons on a bitmap. In addition, when calculating and storing the polygons and associated bitmaps, an angular direction for the laser tracks (cf. DE 4209933 C2) can also be specified. The laser tracks do not necessarily have to follow the raster lines of the bitmap, but methods of computer graphics can be used to calculate the brightness values for a laser track running obliquely to the raster lines, using antialiasing algorithms. (Compare a line running diagonally on a computer screen). When machining the workpiece, a laser device must be used in which the scanner in which the galvanomirrors are located has sufficient mobility with respect to the workpiece in order to be able to move to a position that is as perpendicular as possible to every polygon at a distance of Focal length of the laser optics is located, that is, corresponds to the position that was used as the basis for the calculation of the polygons. For the control of the laser device in the sense of an economic processing, it is necessary to order the polygons in the data set so that they are read by the control electronics in a sequence that has the shortest possible travel of the scanner.

Another object is to avoid dividing lines that arise in the area in which one laser track ends and the next begins (see. DE10032981 A1). There is a particular danger at the edges where two polygons meet.

This object is achieved in that the layer thickness is reduced so much that the resulting boundary line is negligible in height compared to the total height of the scar and is therefore no longer visible. The addition of the dividing line error at the polygon edges is avoided by assigning a separate, independent 3-dimensional polygon mesh to each layer to be removed. This can be chosen completely freely, taking into account the above requirements. It should also be noted that polygon borders overlap (this is inevitable), but must not be on top of each other. Otherwise the dividing line error is added. This means, when considering any point on the surface of the workpiece to be machined and removing material in n layers, that this point "belongs" to n different polygons from n different polygon meshes.

[0027] These polygons can either share a texture bitmap, or else they can be distributed over more than one to a maximum of n bitmaps. With regard to the associated texture bitmaps, it must be noted that if there are several bitmaps, the corresponding layer removal is distributed over the individual bitmaps. This means that the final material removal at a certain point results from an addition of the individual gray values of the texture bitmaps at this point. If the dividing line error can be reduced even more, a method according to DE10032981A can be used, in which an overlap area is formed between the processing sections, in which the processing tracks of the laser interlock in the sections, and the transition points are statistically distributed.

Fig. 1 Schematic representation of the process flow The sequence of the method is shown schematically in the single figure. The method for multi-layer material removal from a workpiece (15) with an arbitrarily shaped three-dimensional surface (1) is carried out by means of a removal means (9) acting point-wise on a surface, such as a laser, by means of which a surface structure (2) on the three-dimensional surface (1 ) is produced. Machining areas (10) are defined on the surface (1), such a machining area (10) being determined by the focus area (11) of the removal means. The surface (1) is approximated by superimposed polygon meshes (18), each polygon (19) of the polygon mesh (18) being assigned to the processing area (10) of the removal means (9). The surface structure (2) is described by at least one grayscale bitmap (14). The grayscale bitmap (14) comprises image points of different grayscale (12) or different color levels. The brightness of the gray level (12) corresponding to each pixel of the gray level bitmap (14) or the intensity of the color level or the characteristic value of the color, such as a wavelength when using multicolored bitmaps, determines the depth of the material removal. The material is removed in the number of layers (7) which correspond to the value of the gray level (12). Each layer (7) is assigned its own polygon mesh (18). The polygon (19) of each layer (7) to be machined does not have an edge section in common with one of the previously machined polygons, so that edge effects can be avoided, which can become visible on the surface by attaching and removing the removal agent. To carry out the method, an original three-dimensional computer model (16) of the workpiece (15) is generated, which is described by an original polygon mesh (17). The three-dimensional corners of the polygons of the original polygon network (17) correspond to two-dimensional points in one or more original texture bitmaps (3). The polygons are transferred into the two-dimensional space of the original texture bitmaps (3), the grayscale value (5) of a pixel (4) of the original texture bitmap (3) corresponding to the required material removal on the workpiece (15) and the processing areas (10) individually Layers (7) comprise. The sum of the machining areas (10) gives the surface (1) and the sum of the layers (7) gives the surface structure (2) of the workpiece (15). Each layer (7) can be written on by a polygon mesh (18), polygon meshes lying one above the other being offset from one another. The surface structure (2) of the workpiece (15) is formed by stacked, mutually offset, Polygon meshes (18) approximated. A grayscale bitmap (14) from a parallel projection of the original texture bitmap (3) onto the polygon (19) within the processing area (10) is assigned to each polygon (19) of the polygon network (18) within the processing area (10) so that the material is removed the removal means (9) can be carried out in each layer (7) according to the values of the gray scale bitmaps (14). The distance value (6) between two layers (7) thus corresponds to the difference in brightness between two adjacent gray levels (12). The original model is derived from the description of the workpiece by means of CAD (spline) surfaces, which represent the original polygon mesh (17). result. The brightness values of the gray levels (12) of the gray level bitmaps (14) are calculated back to the original texture bitmap (3) before or during the processing of the surface (1) of the workpiece (15). Instead of brightness values of gray levels (12), color levels or colors from the color spectrum can also be used.

[0036] List of reference symbols

I. Surface 2. Texture = surface structure

3rd partial area = original texture bitmap

4th pixel

5th grayscale

6th distance value 7th layer

8. free

9. Means of removal

10. Machining area

I I. Focus area 12. Grayscale

13. free

14. Raster image = grayscale bitmap

15. Workpiece

16. Original 3d computer model 17. Original polygon mesh

18. Polygon mesh of a layer

19.Polygon of the mesh 18

Claims

EXPECTATIONS

1. A method for multi-layer material removal from a workpiece (15) with an arbitrarily shaped three-dimensional surface (1) by means of an ablation means (9) acting point-wise on a surface, such as a laser, by means of which a surface structure (2) on the three-dimensional surface (1 ) is generated, with machining areas (10) being defined on the surface (1) and such a machining area (10) being determined by the focus area (11) of the removal means, characterized in that the surface (1) is formed by polygon meshes (18 ) is approximated, each polygon (19) of the polygon network (18) being assigned to the processing area (10) of the removal means (9).

2. The method according to claim 1, characterized in that the surface structure (2) is described by at least one grayscale bitmap (14).

3. The method according to claim 2, characterized in that the gray level bitmap (14) comprises pixels of different gray levels (12) or different color levels.

4. The method according to claim 3, characterized in that the brightness of the gray level (12) corresponding to each pixel of the gray level bitmap (14) or the intensity of the color level determines the depth of the material removal.

5. The method according to claim 4, characterized in that the material removal takes place in the number of layers (7) which correspond to the value of the gray level (12).

6. The method according to claim 1, characterized in that each layer (7) is assigned its own polygon network (18).

7. The method according to claim 7, characterized in that the respective polygon to be processed (19) of each layer (7) has no edge section in common with one of the previously processed polygons.

8. A method for multi-layer material removal from an arbitrarily shaped three-dimensional surface according to claim 1, wherein an original three-dimensional computer model (16) of the workpiece (15) is generated, which is described by an original polygon mesh (17), the three-dimensional corners of the polygons original polygon network (17) correspond to two-dimensional points in one or more original texture bitmaps (3), the polygons being transferred into the two-dimensional space of the original texture bitmaps (3), the grayscale value (5) of a pixel (4) being the original Texture bitmap (3) corresponds to the required material removal on the workpiece (15) and the processing areas (10) comprise individual layers (7), the total of the processing areas (10) the surface (1) and the total of the layers (7) the surface structure ( 2) of the workpiece (15), and each layer (7) can be written on by a polygon mesh (18), and the ob Surface structure (2) of the workpiece (15) is approximated by superposed, mutually offset polygon networks (18), with each polygon (19) of the polygon network (18) within the processing area (10) having a grayscale bitmap (14) from a parallel projection of the Original texture bitmap (3) is assigned to the polygon (19) within the processing area (10) so that the material can be removed by the removal means (9) in each layer (7) according to the values of the grayscale bitmaps (14).

9. The method according to claim 8, characterized in that the master model is derived from the description of the workpiece by CAD (spline) surfaces, which result in the original polygon mesh (17).

10. The method according to claim 8, characterized in that the brightness values of the gray levels (12) of the gray level bitmaps (14) are calculated back to the original texture bitmap (3) before or during the processing of the surface (1) of the workpiece (15).